Cache Me, Catch You: Exploiting LLM Caching Layers in vLLM, GPTCache & Friends
Engineers explore architecture and security mechanisms for Cache Me, Catch You: Exploiting LLM Caching Layers in vLLM, GPTCache & Friends.
Summary
Black Hat session on Cache Me, Catch You: Exploiting LLM Caching Layers in vLLM, GPTCache & Friends.
Related talks
HITCON·42 min
Security researchers examine technical aspects of Prompt Injections in the Wild - Exploiting Vulnerabilities in LLM Agents | HITCON CMT 2023.
The Linux Foundation·41 min
Scaling LLM Inference With Tiered Caching: Extending LMCache With Amazon... Yihua Cheng & Ziwen Ning
Reviews scaling LLM Inference With Tiered Caching: Extending LMCache With Amazon... Yihua Cheng & Ziwen Ning and evaluates key implications for security practitioners.
The Linux Foundation·41 min
Tejas Chopra
Explores headroom: A Context Optimization Layer for LLM Applications and evaluates key implications for security practitioners.